Lethbridge

Advanced
Abstract
Algebra
Dave Witte Morris
University of Lethbridge, Canada

September 4, 2018

To the extent possible under law, Dave Witte Morris has waived
all copyright and related or neighboring rights to this work.

You can copy, modify, and distribute this work, even for commercial purposes, all without asking
permission. For more information, visit
http://creativecommons.org/publicdomain/zero/1.0/
 Contents
Part I. Group Theory v

Chapter 1. Summary of undergraduate group theory 1

1.1. Definitions and examples 1
1.2. Burnside’s Counting Lemma (optional) 7
1.3. Subgroups, conjugates, cosets, and quotient groups 10
1.4. Homomorphisms and isomorphisms 13

Chapter 2. Group Actions 17

2.1. Definition and basic facts 17
2.2. Orbits and stabilizers 18
2.3. Sylow Theorems 21

Index 25

iii
 Part I

Group Theory
 Chapter 1

Summary of
undergraduate group theory

Although this course assumes familiarity with the topics in a typical undergraduate course on
abstract algebra, including subgroups, normal subgroups, homomorphisms, quotient groups, etc.,
we will start with a quick review.

(1.0.1) Notation.
• G is always a group, and X is a set.
• The cardinality of X is denoted |X| (or, sometimes, #X). (Recall that the cardinality of a
set is the number of elements in the set.)
• Z = {. . . , −3, −2, −1, 0, 1, 2, 3, . . .}.
• N = {0, 1, 2, . . .} (unlike some other authors, we include 0 in this set).
• Z+ = N+ = {1, 2, 3, . . .}.
• For k, n ∈ Z, we write k | n to denote that k is a divisor of n (or, equivalently, that n is a
multiple of k).
• For a, b, n ∈ Z, we write a ≡ b (mod n) to denote that n | (a − b), and we may say that a
is congruent to b, modulo n.

§1.1. Definitions and examples

(1.1.1) Definition. A group is a set G together with a binary operation ∗ that is associative and
has an identity element and inverses. (We usually write gh, or sometimes g · h, for g ∗ h.)
• (associative) ∀g1 , g2 , g3 ∈ G, g1 (g2 g3 ) = (g1 g2 )g3.
• (identity element) ∃e ∈ G, ∀g ∈ G, eg = ge = g.
It is easy to show that the identity element of G is unique (see Example 1.1.2(1)). It
is usually denoted by 1 or, sometimes, e. (However, it is denoted by 0 if G is written
additively, which means that the group operation is +.) To avoid confusion, we will
sometimes use 1G for the identity element of G (and 1H for the identity element of
some other group H).
• (inverses) ∀g ∈ G, ∃h ∈ G, gh = hg = 1.
It is easy to show that the inverse of g is unique (see Example 1.1.2(2)). It is denoted
g −1 (unless G is written additively, in which case the inverse is −g) .

1
2 1. SUMMARY OF UNDERGRADUATE GROUP THEORY

(1.1.2) Example. We verify two facts stated in Definition 1.1.1, and also establish two facts of
high-school algebra.
1) If e1 and e2 are identity elements of G, then e1 = e2. To see this, note that:
e1 = e1 e2 (e2 is an identity element, so ge2 = g for all g ∈ G)
= e2 (e1 is an identity element, so e1 g = g for all g ∈ G).
2) Assume that G has an identity element 1, and let g ∈ G. If h1 and h2 are inverses of g, then
h1 = h2. To see this, note that:
h1 = h1 ∗ 1 (1 is the identity element, so h ∗ 1 = h for all h ∈ G)
= h1 ∗ (g ∗ h2 ) (h2 is an inverse of g)
= (h1 ∗ g) ∗ h2 (∗ is associative)
= 1 ∗ h2 (h1 is an inverse of g)
= h2 (1 is the identity element, so 1 ∗ h = h for all h ∈ G).
3) The inverse of a product is the product of the inverses in the reverse order:
(gh)−1 = h−1 g −1 .
To see this, note that
(gh)(h−1 g −1 ) = g(hh−1 )g −1 = g · 1 · g −1 = gg −1 = 1.
A similar calculation shows that (h−1 g −1 )(gh) = 1. Thus, we have shown that if you
multiply gh by h−1 g −1 on either side, then the result is the identity element. This is exactly
what it means to say that h−1 g −1 is the inverse of gh.
4) G has both right and left cancellation: if ag = ah or ga = ha, then g = h. To see this,
note that if ag = ah, then a−1 (ag) = a−1 (ah). Furthermore (using the associative law) the
left-hand side is 1 · g = g and the right-hand side is 1 · h = h, so we conclude that g = h,
as desired. Similarly, if ga = ha, then
g = g · 1 = (ga)a−1 = (ha)a−1 = h · 1 = h.

(1.1.3) Definitions. Let g, h, and a be elements of a group G.

1) The cardinality of the set G is called the order of G. It is denoted |G|.
2) For k ∈ Z+, we define g k to be the product of k copes of g: we have g k = gg · · · g, where
there are k factors on the right-hand side. And we let g −k = (g k )−1 (or (g −1 )k, which is the
same thing, by Example 1.1.2(3)). Finally, we let g 0 = 1. With these definitions, the usual
laws of exponents hold (for k, ℓ ∈ Z):
g 0 = 1, g 1 = g, g k g ℓ = g k+ℓ , (g k )ℓ = g kℓ , (g k )−1 = (g −1 )k .
(If the group operation is +, then we write kg for g + g + · · · + g, instead of g k.)
3) The order of g is the smallest k ∈ Z+, such that g k = 1. It is denoted |g|. (If no such k
exists, then |g| = ∞.)
4) g and h commute if gh = hg. (We may also say that g centralizes h.) We say that G is
abelian (or commutative) if all the elements of G commute with each other.

(1.1.4) Exercise. Let g ∈ G, such that g has finite order.

1) For k, ℓ ∈ Z, show:
(a) g k = 1 if and only if k is a multiple of |g|.
(b) More generally, g k = g ℓ a k ≡ ℓ (mod |g|).
2) Show |g −1 | = |g|.

In this course, we will mostly be interested in finite groups. (These are groups that have only
finitely many elements, or in other words, the groups whose order is finite.) It is important to be
familiar with some examples.
 1.1. DEFINITIONS AND EXAMPLES 3

(1.1.5) Example (Integers modulo n under addition). Recall that Zn is the set of integers modulo n.
This means that elements of Zn are integers, except that we consider two elements of Zn to be
equal if they are congruent modulo n. More precisely, the elements of Zn are equivalence classes
of integers, where two integers are equivalent if they have the same remainder when you divide
them by n. We can use k to represent the equivalence class of k in Zn. For example,
Z2 = {0, 1}, Z3 = {0, 1, 2}, Z4 = {0, 1, 2, 3}, Z5 = {0, 1, 2, 3, 4}, Z6 = {0, 1, 2, 3, 4, 5}, etc.
Each of these is a group under addition modulo n, which is defined by k + ℓ = k + ℓ:
1) (associative) In elementary school, we all learned the associative law (for the addition of
ordinary integers). This implies that addition modulo n is also associative:
( )
k+ℓ +m =k+ℓ+m (definition of addition in Zn )
= (k + ℓ) + m (definition of addition in Zn )
= k + (ℓ + m) (associate law for addition in Z)
=k+ℓ+m (definition of addition in Zn )
=k+ℓ+m (definition of addition in Zn ).
2) (identity element) The additive identity element of Z is 0, so the identity element of Zn is 0:
k+0=k+0=k and 0 + k = 0 + k = k.
3) (inverses) The (additive) inverse of k in Z is −k so the inverse of k in Zn is −k:
k + −k = k + (−k) = 0 and −k + k = (−k) + k = 0.
Note that, for the element 1 ∈ Zn, we have |1| = n.
(1.1.6) Exercise. The direct product of two groups G and H is the Cartesian product G × H with
componentwise multiplication. That is, (g1 , h1 ) ∗ (g2 , h2 ) = (g1 g2 , h1 h2 ).
1) Show that G × H is a group.
[Hint: By definition, this means you need to show that the operation is associative, and has an identity element
and inverses.]
2) For (g, h) ∈ G × H, show that |(g, h)| is the least common multiple of |g| and |h|.
(1.1.7) Example. Z3 × Z5 × Z7 is a group of order 3 · 5 · 7 = 105. In this group, we have
(2, 4, 5) + (2, 1, 4) = (2 + 2, 4 + 1, 5 + 4) = (2 + 2, 4 + 1, 5 + 4) = (4, 5, 9) = (1, 0, 2).
(1.1.8) Exercise (semidirect product of cyclic groups). Suppose m, n ∈ Z+, and let k ∈ Z, such that
km ≡ 1 (mod n). Define a binary operation on the set Zm × Zn by
(x1 , y1 ) ∗ (x2 , y2 ) = (x1 + ky1 x2 , y1 + y2 ).
Show that this operation is associative, and has an identity element and inverses. Hence, it defines
a group (of order mn).
This group is called the semidirect product of Zm and Zn (with multiplier k), and is denoted
Zm ⋊k Zn. It is a generalization of the the direct product of Zm and Zn, which is the special case
where k = 1. (The construction can be generalized by replacing Zm and Zn with other groups, but
we will not discuss this.)
(1.1.9) Example. In Z7 ×4 Z3, we have
(3, 2) ∗ (5, 1) = (3 + 42 · 5, 2 + 1) = (3 + 42 · 5, 2 + 1) = (83, 3) = (6, 0).
Groups often arise as symmetries of an object.
(1.1.10) Definition (informal). A symmetry of an object is a way of repositioning the object in
such a way that it occupies exactly the same space as it did originally. It is sometimes called an
“undetectable motion:” if you perform a symmetry to an object while someone is not looking, they
will not realize that anything has changed, because everything looks exactly as it did before.
4 1. SUMMARY OF UNDERGRADUATE GROUP THEORY

(1.1.11) Example (Rotations of a square). Imagine a square lying on a tabletop. Rotating the square
by 90◦ is a symmetry of the square. For short, let us use rθ to denote a rotation by θ degrees
(clockwise). So r90 is a symmetry of the square. Other symmetries are r180 and r270. (By the
way, another name for r270 is r−90. Or we could rotate by 0◦ (doing nothing certainly leaves the
square occupying the same space as it did before), so r0 = r360 is also a symmetry of the square.
There are no other rotational symmetries, so the rotational symmetries of the square form the set
{r0 , r90 , r180 , r270 }.
It is important to note that this set is a group under composition, or, in the language of non-
mathematicians, the “after” operation: recall that r ◦ s is the procedure that is obtained by doing
r after s: first apply s to the object, then apply s. For example, r90 ◦ r180 = r270, because rotating
the square by 90 degrees after rotating it by 180 degrees has exactly the same effect as applying a
single rotation of 180 degrees to the square. Note that this is a group of order 4. We have |r0 | = 1,
|r90 | = |r270 | = 4, and |r180 | = 2.

In general, the symmetries of an object always form a group.

(1.1.12) Remark. It is not difficult to verify the above statement that symmetries form a group:
• To see that composition is a binary operation on the set of symmetries, we need to know
that the set of symmetries is closed under composition: if r and s are symmetries of
an object X, then r ◦ s is also a symmetry of X. Fortunately, this is clearly true: if we
reposition X in some way that is undetectable, and then reposition it again in a way that is
undetectable, then the final position is also indistinguishable from the original position.
• Composition is associative: r ◦(s ◦t) = (r ◦s)◦t. This is because it does not matter whether
we :
◦ first do t, then do s, and then take a break before doing r, or
◦ first do t, then take a break before doing s and then r.
In either case, we are doing t, and then s, and then r.
• The identity element of the group is the “do nothing” operation.
• The inverse of a particular symmetry is “put it back the way it was.”

The square can be replaced with any regular polygon:

(1.1.13) Example. The set of rotations of a regular n-gon P is a group Rot(P ) of order n. More
precisely, if θ = 360/n, then
Rot(P ) = {r0 , rθ , r2θ , . . . , r(n−1)θ }.
−1
Furthermore, we have rkθ ◦ rℓθ = r(k+ℓ)θ, and rkθ = r−kθ.

(1.1.14) Exercise. Find the order of each of the six rotations of the regular hexagon.

But rotations are not the only symmetries of a regular n-gon:

(1.1.15) Example (dihedral group). In addition to rotational symmetry, a

regular n-gon also has reflection symmetry: it looks the same in a mirror.
Another way of looking at this is that the square can be “flipped over.” For
example, if S is a square in the xy-plane, with its centre at the origin and
its sides parallel to the coordinate axes, then rotating S by 180◦ about the
x-axis is a symmetry of S. (Another way of describing this is to perform a
mirror symmetry across the x-axis: we are simply moving each point (x, y)
to (x, −y).) The y-axis is also an axis of symmetry. And there are two other
reflection axes that go through opposite corners of the square. They are
illustrated in the figure at right.
In general, a regular n-gon has precisely n reflection symmetries: the axes are lines through
the origin whose intersections with the boundary of the n-gon are either at a corner or are the
exact midpoint of an edge.
 1.1. DEFINITIONS AND EXAMPLES 5

Thus, the full group of symmetries of a regular n-gon has order 2n: there are n rotations and
n reflections. (Every reflection has order 2, but the orders of the rotations are various numbers
between 1 and n, inclusive.) This group is called the dihedral group of order 2n, and is denoted
D2n. For example, the group of symmetries of a square is D8, and the group of symmetries of a
regular pentagon is D10.
(1.1.16) Other conventions. Some textbooks use the notation Dn for the dihedral group of order 2n,
instead of D2n.
We can also look at objects in 3-dimensions, instead of the plane:
(1.1.17) Example (rotations of a cube). A cube has 24 rotational symmetries. To see this, consider
the cube to being resting on a tabletop. We can move any of the 6 faces of the cube to be on the
bottom. After this, the face that is now on the bottom can be in any of 4 positions (because the
face is a square, and a square has 4 rotational symmetries). Therefore, we have 6 choices for which
face to put on the bottom, and 4 choices of position after that. (And these two choices completely
determine the position of the cube.) So the total number of possible positions (or symmetries) is
6 × 4 = 24.
(1.1.18) Exercises.
1) Recall that the Platonic solids are:
(a) regular tetrahedron (the 4 faces are equilateral triangles),
(b) cube (the 6 faces are squares),
(c) regular octahedron (the 8 faces are equilateral triangles),
(d) regular dodecahedron (the 12 faces are regular pentagons), and
(e) regular icosahedron (the 20 faces are equilateral triangles).
What is the order of the group of rotations of each of these solids?
[Hint: Let us say that the solid is resting on a tabletop. You may assume, without proof, that any face can be
rotated to be on the bottom, that each rotational symmetry of the bottom face extends to a symmetry of the
entire solid, and that these two choices determine the symmetry of the solid. (This is because we are dealing only
with Platonic solids.)]
2) How many symmetries does a rectangle have if it is not a square? (Include both rotations
and reflections.)
3) Describe each of the rotational symmetries of a regular tetrahedron. In particular, how
many are there of each order?
4) Describe each of the 24 rotational symmetries of a cube. In particular, how many are there
of each order?
We can also look at symmetries of a collection of objects, instead of a single object. For
example, if there are several identical items sitting on a table, then we could interchange some of
them, and the scene would look exactly the same as before. This is an example of a permutation.
To keep track of what we are doing, it is helpful to number the items, so that we can say, for
example, that we interchanged item #2 with item #6.
(1.1.19) Definition.
1) Recall that a function f : X → Y is:
• one-to-one (or injective) iff for all x1 , x2 ∈ X, such that f (x1 ) = f (x2 ), we have
x1 = x2;
6 1. SUMMARY OF UNDERGRADUATE GROUP THEORY

• onto (or surjective) iff for all y ∈ Y, there exists x ∈ X, such that f (x) = y;
• a bijection iff f is both one-to-one and onto.
2) A permutation of X is a bijection from X to itself.
3) The set of all permutations of X is called the symmetric ( ) group on X. It is a group under
composition (see Exercise 1.1.24): (σ τ)(x) = σ τ(x) for all x ∈ X.
4) The symmetric group on {1, 2, . . . , n} is called the symmetric group on n letters (or “of
degree n”), and is denoted Sn. Its order is n!.
5) For x1 , . . . , xk ∈ X, we use (x1 x2 . . . xk ) to denote the unique permutation σ ∈ Sn, such
that
• σ (xi ) = xi+1 for i ∈ {1, . . . , k} (reading the subscript modulo k), and
• σ (x) = x for all x ∉ {x1 , x2 , . . . , xk }.
Such a permutation is called a cycle of length k, or a k-cycle.
6) Two cycles (x1 x2 . . . xk ) and (y1 y2 . . . yℓ ) are disjoint if the sets {x1 , x2 , . . . , xk } and
{y1 , y2 , . . . , yℓ } are disjoint (that is, they have no elements in common).
Dealing with permutations requires some basic facts from Math 2000. If you have difficulty
with the definition, or the subsequent exercises, you may want to review a textbook for that course.
One such textbook is available online at:
http://people.uleth.ca/~dave.morris/books/proofs+concepts.html
(1.1.20) Definition. Assume φ : X → Y and σ : Y → Z. The composition of σ and φ is the function
σ ◦ φ from X to Z that is defined by
( )
(σ ◦ φ)(x) = σ φ(x) for all x ∈ X.
(1.1.21) Exercises (Basic properties of composition). Assume φ : X → Y and σ : Y → Z.
1) Prove that composition is associative: show that if, in addition to φ : X → Y and σ : Y → Z,
we also have τ : Z → W, then (τ ◦ σ ) ◦ φ = τ ◦ (σ ◦ φ). (This was explained informally in
Remark 1.1.12, but you should be able to write an official proof of this fact.)
2) Show that the composition of one-to-one functions is one-to-one: if φ and σ are one-to-one,
then σ ◦ φ is one-to-one.
3) Show that the composition of onto functions is onto: if φ and σ are onto, then σ ◦ φ is
onto.
4) Show that the composition of bijections is a bijection: if φ and σ are bijections, then σ ◦ φ
is a bijection.
[Hint: Use (2) and (3).]

Bijections arise in many situations, but one of their most important applications is in showing
that two sets have the same cardinality:
(1.1.22) Basic facts.
1) Definition. Two sets X and Y have the same cardinality if and only if there is a bijection
from X to Y.
2) Assume φ : X → Y, and let A be any subset of X.
(a) Definition. f (A) = { f (a) | a ∈ A }. (This is called the image of A under f.)
(b) If φ is one-to-one, then |f (A)| = |A|.
It is not difficult to see if φ : X → Y has an inverse φ−1 : Y → X, then φ must be a bijection.
It is an important fact from Math 2000 that the converse is true. However, this is more difficult,
so you do not need to try to prove it for yourself, although you should remember this important
fact:
(1.1.23) Basic fact. If φ : X → Y is a bijection, then φ has an inverse φ−1 : Y → X.
Some of the above facts will be helpful in solving the following problem:
 1.2. BURNSIDE’S COUNTING LEMMA (optional) 7

(1.1.24) Exercise. Prove that the symmetric group SX is indeed a group (under composition).
Also show that the identity element of this group is the identity map on X. (The identity map
on X is the function 1X : X → X defined by 1X (x) = x for x ∈ X.)
(1.1.25) Basic facts.
1) If two cycles are disjoint, then they commute with each other.
2) Every permutation of a finite set is a product of disjoint cycles. Furthermore, this
decomposition into disjoint cycles is unique, up to a permutation of the factors.
(1.1.26) Example. (1 3)(2 5 7) is an element of S7. (In fact, it is an element of Sn for any n ≥ 7). Its
action on the elements of {1, 2, 3, 4, 5, 6, 7} is:
1 , 3, 2 , 5, 3 , 1, 4 , 4, 5 , 7, 6 , 6, 7 , 2.
This permutation has order 6. (In general, the order of a permutation is the least common multiple
of the lengths of the cycles in its decomposition as a product of disjoint cycles.)
(1.1.27) Exercise. In each part of the problem, write each of the given permutations as a product
of disjoint cycles. Also find the order of each of the permutations.
1) (1 2)(2 3)(3 4). 2) (1 2)(1 3)(1 4)(1 5). 3) (2 3 5)(2 3 4)(1 4 2 5 3).
4) Number the corners of a regular hexagon from 1 to 6, clockwise, so that we may think of
each of its six rotational symmetries as elements of S6. You are to consider all six of these
permutations.
5) Number the corners of a square from 1 to 4, clockwise, so that we may think of each of
the elements of D8 as elements of S6. The permutations to consider are:
(a) the reflection whose axis is through corners 1 and 3, and
(b) the reflection whose axis is through the midpoints of sides 1−2 and 3−4.

§1.2. Burnside’s Counting Lemma (optional)

(1.2.1) Problem. We have five colours of paint available with which to paint the faces of a cube.
(Every face needs to be painted, and only one colour is allowed to appear on each face, but several
faces may be painted the same colour, and it is not necessary to use all 5 colours.) How many
essentially different ways are there to paint the cube?
High-school-level mathematics reveals there are exactly 56 = 15, 625 ways to paint the faces
of the cube, because we can use any one of 5 colours of paint on each of the 6 faces of the cube.
However, this is not the answer we want. For example, suppose Alice paints the top face of her
cube blue and all of the other faces red, while Bob paints the bottom of his cube blue and all the
other faces red. Then Alice and Bob have painted their cubes in essentially the same way — if
Alice turns her cube upside-down, then it looks just like Bob’s.
If there were only two colours of paint, it would not be difficult to solve the problem by listing
all of the possible colourings. But that is not very feasible for five colours (unless a computer is
used). This section presents a method that easily solves this problem and many related types of
problems, by applying the theory of group actions.
We will see the official definitions in Section 2.1, but, for now, it will suffice to have an informal
understanding of two key ideas:
• Saying that a group G acts on a set X means that each element g of G acts like a permutation
of the elements of a set X: the group element g carries each x ∈ X to some point that is
called gx.
• Each x in X can be moved to certain other places in X. (But probably there are places that
G cannot move it to. For example, rotations of a square cannot move a corner point to a
point that is not on a corner.) The set of points that x can move to is called the orbit of x.
(1.2.2) Definition. Suppose G acts on X. For g ∈ G and x ∈ X, we say that x is a fixed point of g
if gx = x. (This means that g leaves x alone, instead of moving it somewhere else.)
8 1. SUMMARY OF UNDERGRADUATE GROUP THEORY

(1.2.3) Proposition (Burnside’s Counting Lemma). Suppose G acts on a finite set X (and G is finite).
Then the number of orbits of G on X is equal to the number of fixed points of the average element
of G.
(1.2.4) Remark. To be precise, the phrase “the number of fixed points of the average element
of G,” actually means “the average of the numbers of fixed points of elements of G.” That is, list the
number of fixed points of each element of G, and then calculate the average (or arithmetic mean)
of these numbers.
We will see some applications of this proposition right now, but the proof will be postponed
until Section 2.2, when we have more tools from the theory of group actions.
(1.2.5) Examples.
1) Suppose every element of G leaves all of the points of X alone. (This could be called the
“lazy” group action, because the elements of G are too lazy to do anything, but it is officially
called the trivial action of G.) Then each element of X constitutes an orbit, so the number
of orbits is |X|. On the other hand, each element of G fixes every point in X, so the number
of fixed points of each element (including the average element) is also |X|. This agrees with
the conclusion of Burnside’s Counting Lemma.
2) Let G = Rot(cube) and let X be the set of faces of the cube. Only the identity element of G
fixes every face of the cube, while there are 9 rotations that fix two faces (six rotations of
order 4, and three rotations of order 2), and all other elements of G have no fixed points
in the action on X. So the number of fixed points of the average element of G is
1 × 6 + 9 × 2 + 14 × 0
= 1.
24
Also, any face of the cube can be moved to any other face by a rotation, so this action
has only 1 orbit. Hence, we once again have agreement with the conclusion of Burnside’s
Counting Lemma.
3) Let G be the group of symmetries of a rectangle (which is not a square), and let X be the set
of sides of the rectangle. The rectangle has only four symmetries (see Exercise 1.1.18(2)):
• The trivial symmetry (do nothing) fixes all 4 sides.
• r180 does not fix any sides, so the number of sides that are fixed is 0.
• Each of the two reflection symmetries fixes the 2 sides that are on its axis (and
interchanges the other two).
So the number of fixed points of the average element of G is
1×4+1×0+2×2
= 2.
4
Also the long sides of the rectangle can be moved to each other, and the short sides of
the rectangle can be moved to each other, but a symmetry cannot move a long side so a
short side. So there are 2 orbits. Therefore, of course, we yet again have agreement with
the conclusion of Burnside’s Counting Lemma.
Solution to Problem 1.2.1. Two colourings of the cube are essentially different if there is no
rotation of the cube that takes one to the other, that is, if they are in different orbits in the action
of Rot(cube) on the collection of colourings of the cube. So the total number of essentially different
colourings of the cube is equal to the number of orbits of Rot(cube) on the set of colourings. By
Burnside’s Counting Lemma, this is the same as the number of colourings fixed by an average
element of Rot(cube).
Any permutation can be uniquely expressed as a product of disjoint cycles, and it is easy to
see that such a permutation will fix a colouring if and only if each cycle of the permutation is
monochromatic. (That is, all of the faces in each cycle must have the same colour, but faces in
different cycles may have different colours.) Thus, if there are k colours available, then the number
of colourings fixed by the permutation g is kcyc(g), where cyc(g) is the number of cycles in the
disjoint cycle decomposition of g. Therefore, applying Burnside’s Counting Lemma tells us:
 1.2. BURNSIDE’S COUNTING LEMMA (optional) 9

The number of essentially different ways to paint the faces of a cube with k colours is
1 ∑
kcyc(g) .
|Rot(cube)| g∈Rot(cube)
To calculate the sum, we count the cycles in each element of Rot(cube):
1) The identity element fixes all 6 faces, so it has 6 cycles of length one.
2) 3 elements are rotations of 180◦ about an axis through opposite faces of the cube. These
have 2 cycles of length one, and 2 cycles of length two, for a total of 4 cycles.
3) 6 elements are 90◦ rotations. They have 2 cycles of length one, and 1 cycle of length four,
for a total of 3 cycles.
4) 6 elements are 180◦ rotations about an axis through opposite edges. They have 3 cycles,
all of length two.
5) 8 elements are rotations about an axis through opposite corners. They have 2 cycles of
length three.
So the number of essentially different colourings is
k6 + 3k4 + 6k3 + 6k3 + 8k2
. (1.2.6)
24
We plug in k = 5 to get 800 as the answer to the original problem. □

(1.2.7) Remark. Letting k = 1 in (1.2.6) yields the value 1. This agrees with the observation that
there is only one way to colour the faces of a cube if only one colour of paint is available. (The
entire cube must be covered with that one colour.)
(1.2.8) Exercises.
1) Letting k = 2 in (1.2.6) tells us that there are exactly 10 essentially different ways to colour
the faces of a cube with two colours. Describe (or draw) each of these 10 colourings.
2) How many essentially different ways are there to colour the faces of a cube with three
colours?
3) How many essentially different ways are there to colour the vertices of an equilateral
triangle if k colours of paint are available?
(a) Assume two colourings are equivalent if one can be obtained from the other by rotating
the triangle.
For k = 3, there are 11 different colourings:

(b) Assume two colourings are equivalent if one can be obtained from the other by
applying any symmetry of the triangle (that is, either a rotation or a reflection of
the triangle).
For k = 3, there are 10 different colourings. They are the same colourings pictured
above, except that there is now only one colouring that uses all 3 colours.
4) Replace the triangle in Exercise 3 with a regular p-gon, where p is a prime number (and
p ≥ 3). Consider both (a) rotations only and (b) all of the symmetries in the dihedral
group D2p.
5) For the special case of Exercise 4a with p = 5 and k = 2, draw one colouring from each of
the 8 different equivalence classes.
6) Replace the cube in Problem 1.2.1 with a regular tetrahedron (and find a formula that
solves this problem for any number k of colours, not only for 5 colours).
10 1. SUMMARY OF UNDERGRADUATE GROUP THEORY

§1.3. Subgroups, conjugates, cosets, and quotient groups

(1.3.1) Definitions (subgroups).
1) A subset H of G is a subgroup of G if it is closed under the group operations: for all
h1 , h2 , h ∈ H, we have h1 h2 ∈ H and h−1 ∈ H (and 1 ∈ H). Equivalently, this means that
H is itself a group under the operation obtained by restricting the operation of G.
(For example, if P is a regular n-gon, then Rot(P ) is a subgroup of D2n.)
2) The obvious subgroups of G are {1} and G. (There may or may not be other subgroups
of G.)
• {1} is the trivial subgroup of G. (All other subgroups are nontrivial.)
• A subgroup H of G is said to be proper if H ≠ G.
3) For g ∈ G, we let ⟨g⟩ = { g k | k ∈ Z }. This is an abelian subgroup of G (see Exercises 1.3.4(1)
and 1.3.5(1)), and is called the cyclic subgroup generated by g.
4) We say G is cyclic if G = ⟨g⟩ for some g ∈ G. For example, Zn is cyclic (for every n ∈ Z+),
because Zn = ⟨1⟩.
(1.3.2) Example. If H1 and H2 are subgroups of G, then H1 ∩ H2 is a subgroup of G.
Proof. We wish to show that H1 ∩ H2 is closed under multiplication and inverses, and contains the
identity element.
(closed under multiplication) Given g, h ∈ H1 ∩ H2 we have g, h ∈ H1 and g, h ∈ H2. Since
each Hi is a subgroup, and is therefore closed under multiplication, this implies that gh ∈ H1 and
gh ∈ H2. So gh ∈ H1 ∩ H2, as desired.
(closed under multiplication) Given h ∈ H1 ∩ H2 we have h ∈ H1 and h ∈ H2. Since each Hi is
a subgroup, and is therefore closed under inverses, this implies that h−1 ∈ H1 and h−1 ∈ H2. So
h−1 ∈ H1 ∩ H2, as desired.
(identity element) Since each Hi is a subgroup, it must contain the identity element. This means
that 1 ∈ H1 and 1 ∈ H2. So 1 ∈ H1 ∩ H2, as desired. □

(1.3.3) Remark. By induction, Example 1.3.2 implies that the intersection of any finite collection
of subgroups of G is a subgroup of G, because H1 ∩ H2 ∩ · · · ∩ Hn = (H1 ∩ H2 ∩ · · · ∩ Hn−1 ) ∩ Hn
is the intersection of two subgroups. But it is not difficult to show directly that the intersection
of any collection of subgroups is also a subgroup, even if the collection is infinite.
(1.3.4) Exercise. Here are some important subgroups of G:
1) Let g ∈ G. Show that ⟨g⟩ is a subgroup of G.
2) For any subset S of G, the centralizer of S in G is
CG (S) = { g ∈ G | g commutes with every element of S }.
Show that CG (S) is a subgroup of G.
3) For any subgroup H of G, the normalizer of H in G is
NG (H) = { g ∈ G | gH = Hg },
where gH = { gh | h ∈ H } and Hg = { hg | h ∈ H }. Show that NG (H) is a subgroup of G.
4) The centre of G is
Z(G) = CG (G) = { z ∈ G | zg = gz for all g ∈ G }.
Show that this is an abelian, normal subgroup of G.
(1.3.5) Exercises.
1) Show that every cyclic group is abelian.
2) For g ∈ g, show that |g| = |⟨g⟩|. More precisely, show that ⟨g⟩ = { g k | 0 ≤ k < |g| }, and
that all of these elements are distinct.
[Hint: Item 1(1b).]
 1.3. SUBGROUPS, CONJUGATES, COSETS, AND QUOTIENT GROUPS 11

(1.3.6) Exercise. Suppose H and K are subgroups of G, and let HK = { hk | h ∈ H, k ∈ K }. Show

that HK is a subgroup of G if and only if HK = KH.
(1.3.7) Definitions.
1) For any subset S of G, we use ⟨S⟩ to denote the (unique) smallest subgroup of G that
contains S. This is called the subgroup of G generated by S. The elements of ⟨S⟩ are
k k k
precisely the products of the form s1 1 s2 2 · · · sr r , where r ∈ N+ and each si ∈ S and ki ∈ Z
(see Exercise 1.3.8).
2) We say that a subset S of G is a generating set for G (or that S generates G) if ⟨S⟩ = G.
(1.3.8) Exercise. Let S be a nonempty subset of G, and let H be the set of all elements of G that
k k k
can be written as a product of the form s1 1 s2 2 · · · sr r , where r ∈ N+ and each si ∈ S and ki ∈ Z.
Show that H is the unique smallest subgroup of G that contains S. More precisely, show:
1) H is a subgroup of G that contains S, and we have H ⊆ H ′, for every subgroup H ′ of G
that contains S. (This is what it means to say that H is the smallest subgroup of G that
contains S: it is contained in all of the others.)
2) No other subgroup of G satisfies the conditions in (1).
(1.3.9) Definitions (conjugates and normal subgroups).
1) For any g, h ∈ G, we let
g
h = ghg −1 .
This is called the conjugate of H by g. Note that
g
h=h a gh = hg a h commutes with g.
2) For any subgroup H of G and g ∈ G, we let

g
H = { gh | h ∈ H }.
This is called the conjugate of H by g. It is a subgroup of G (see Exercise 1.3.12(2)), and
we have |gH| = |H| (see Exercise 1.4.6(2)).
3) We say that a subgroup K is conjugate to H if K = gH, for some g ∈ G.
4) We say that g normalizes H if gH = H.
5) A subgroup N of G is normal if every element of g normalizes N. When this is the case,
we write N ⊴ G.
(1.3.10) Warning. When g normalizes H, we know that gh is some element of H, for every h ∈ H.
However, there is no reason to expect gh to be equal to h. Usually, conjugation by g will move the
elements of H around to different places. The following is an example.
(1.3.11) Example. The group of rotations is a normal subgroup of D2n. Any two rotations commute,
but it is not difficult to see that if f is a reflection, then frθ = r−θ.
(1.3.12) Exercises.
1) Show that every subgroup of an abelian group is normal: if G is abelian and H is a subgroup
of G, then H ⊴ G.
2) Show that if H is a subgroup of G, and g ∈ G, then gH is a subgroup of G.
3) Let H be a subgroup of G. Show
g
H⊴G a gH = Hg for all g ∈ G a H ⊆ H for all g ∈ G.
4) Show that if N1 and N2 are normal subgroups of G, then N1 ∩ N2 is a normal subgroup
of G.
5) Suppose H and K are subgroups of G. Show that if every element of H normalizes K (or
vice-versa), then HK is a subgroup of G.
12 1. SUMMARY OF UNDERGRADUATE GROUP THEORY

(1.3.13) Remark. Much as in Remark 1.3.3, Exercise 1.3.12(4) implies that the intersection of any
finite collection of normal subgroups of G is a normal subgroup of G. But it is not difficult to show
directly that the intersection of any collection of normal subgroups is also a normal subgroup,
even if the collection is infinite.

(1.3.14) Definitions (cosets). Let H be a subgroup of G, and let g ∈ G.

1) The set gH is called a left coset of H. The collection of all left cosets is denoted G/H, and
it is a partition of G (see Example 1.3.17). (Recall that a partition of G is a collection of
nonempty subsets of G whose union is all of G.)
2) The number of left cosets is the index of H, and is denoted |G : H|.

(1.3.15) Exercises. Let H be a subgroup of G.

1) For all g, g1 , g2 ∈ G, show:
(a) g ∈ gH, (b) gH = H a g ∈ H, (c) g1 H = g2 H a g1−1 g2 ∈ H.
2) Show that all cosets of H have the same cardinality: for all g1 , g2 ∈ G, we have
|g1 H| = |g2 H| = |Hg2 | = |Hg1 | = |H|.
3) Prove Lagrange’s Theorem: If G is finite, and H is a subgroup of G, then |G : H| = |G|/|H|.
Therefore, the order of every subgroup of G is a divisor of the order of G.
4) Prove that every group of prime order is cyclic.
[Hint: Lagrange’s Theorem.]

5) Prove that if |G| is finite, then g |G| = 1 for all g ∈ G.

[Hint: Lagrange’s Theorem implies |g| | |G|.]

6) Show that every subgroup of index 2 is normal: if H is a subgroup of G, and |G : H| = 2,

then H ⊴ G.

(1.3.16) Remark. The converse of Lagrange’s Theorem is not true: there are examples where G has
no subgroup of order n, even though n is a divisor of |G|.

(1.3.17) Example. For any subgroup H of G, we show that the left cosets of H form a partition
of G. In other words, we show that every element of G is in a unique left coset. To see this, first
note that Exercise 1.3.15(1) tells us g ∈ gH. This establishes that each element of g is in a left
coset, so all that remains is the uniqueness. Suppose g ∈ g1 H and g ∈ g2 H. This means that, for
each i, we may write g = gi hi, with hi ∈ H. Therefore gi = gh−1 −1
i . Also note that hi ∈ H (since
hi ∈ H and the subgroup H must be closed under inverses), so we see from Exercise 1.3.15(1) that
h−1
i H = H. Therefore
gi H = (gh−1 −1
i )H = g(hi H) = gH.

So g1 H = gH = g2 H. So any two left cosets of H that contain g are equal. This establishes the
desired uniqueness of the left coset that contains g.

(1.3.18) Remark. The set Hg is a right coset of H. The right cosets also form a partition of G,
and it is not difficult to see that the number of right cosets is equal to the number of left cosets.
Some textbooks develop the theory by using right cosets instead of left cosets; which side to use
is purely a matter of choice. For more advanced topics, it is often convenient to have both right
cosets and left cosets available, but left cosets will suffice for our purposes this semester.

(1.3.19) Definition. If N is a normal subgroup of G, then G/N is a group under the operation
defined by (g1 N)(g2 N) = g1 g2 N (see Exercise 1.3.20(2)). This is called the quotient of G by N.

(1.3.20) Exercises.
1) Prove that quotients of abelian groups are abelian: if G is abelian and N is a subgroup of G,
then G/N is abelian.
 1.4. HOMOMORPHISMS AND ISOMORPHISMS 13

2) Let N be a normal subgroup of G.

(a) Show that the binary operation on G/N is well-defined. That is, show that the product
(g1 N)(g2 N) depends only on the cosets g1 N and g2 N, and not on the particular
representatives g1 and g2.
More precisely, show that if g1 N = g1′ N and g2 N = g2′ N, then g1 g2 N = g1′ g2 N.
[Hint: You definitely need to use the fact that N is normal!]
(b) Show that the binary operation on G/N is associative, and has an identity element and
inverses.

§1.4. Homomorphisms and isomorphisms

(1.4.1) Definition. Assume G and H are groups. A function φ : G → H is a homomorphism if it
respects the group operation. By this, we mean that φ(g1 g2 ) = φ(g1 ) φ(g2 ) for all g1 , g2 ∈ G. More
precisely, if the operations on G and H are ∗ and •, respectively, then φ(g1 ∗ g2 ) = φ(g1 ) • φ(g2 ).
(1.4.2) Example. The set R of real numbers is a group under addition (+), and the set R+ of positive
real numbers is a group under multiplication (·). The exponential map ex is a homomorphism
from R to R×, because it turns addition into multiplication: ex+y = ex · ey .
(1.4.3) Exercises. Assume φ is a homomorphism from G to H.
1) By definition, we know that φ respects multiplication.
(a) Show that φ also respects the identity element and inverses. This means that
φ(1G ) = 1H , and φ(g −1 ) = φ(g)−1 for all g ∈ G.
(b) Show that φ also respects powers: φ(g k ) = φ(g)k, for all g ∈ G and k ∈ Z.
2) Show that if K is a subgroup of G, then φ(K) is a subgroup of H.
(Warning: φ(K) might not be a normal subgroup of H, even if K is a normal subgroup
of G.)
3) We let
ker φ = φ−1 (1H ) = { g ∈ G | φ(g) = 1H }.
This is called the kernel of φ. Show that ker φ is a normal subgroup of G.
4) Conversely, suppose N is any normal subgroup of G. Show that the function ψ(x) = xN
is a homomorphism from G to G/N whose kernel is N.
5) Show that φ is one-to-one if and only if the kernel of φ is trivial.
(1.4.4) Definition.
1) A bijective homomorphism is called an isomorphism.
2) We say that groups G and H are isomorphic (and write G ≊ H) if there is an isomorphism
from G to H. This is an equivalence relation (see Exercise 1.4.7(4)).
Isomorphisms preserve all properties that can be expressed in group-theoretic terms. For
example:
≊
(1.4.5) Exercises. Assume φ : G -→ H, and let g, g ′ ∈ G. Then:
1) |G| = |H|.
2) G is abelian if and only if H is abelian.
3) g commutes with g ′ if and only if φ(g) commutes with φ(g ′ ).
4) |φ(g)| = |g|.
5) G is cyclic if and only if H is cyclic.
.
. etc.
.

(1.4.6) Exercises.
1) Show that every cyclic group of order n is isomorphic to Zn.
14 1. SUMMARY OF UNDERGRADUATE GROUP THEORY

2) Assume H is a subgroup of G, and let g ∈ G. Show H ≊ gH. (By Exercise 1.4.5(1), this
implies |H| = |gH|.)
[Hint: Define φ : H → gH by φ(h) = gh.]

(1.4.7) Exercises.
1) Show that the composition of homomorphisms is a homomorphism. This means that if φ
is a homomorphism from G to H, and σ is a homomorphism from H to K (where G, H,
and K are groups), then the composition σ ◦ φ is a homomorphism from G to K.
2) Show that the composition of isomorphisms is an isomorphism. This means that if φ is
an isomorphism from G to H, and σ is an isomorphism from H to K (where G, H, and K
are groups), then the composition σ ◦ φ is an isomorphism from G to K.
[Hint: Use previous exercises.]
3) Show that the inverse of an isomorphism is an isomorphism: if φ is an isomorphism from G
to H, then φ has an inverse (which is a function from H to G), and φ−1 is an isomorphism
from H to G.
4) Show that isomorphism is an equivalence relation on the collection of all groups. In other
words, show that isomorphism is:
• reflexive: G ≊ G,
• symmetric: G ≊ H ⇒ H ≊ G, and
• transitive: G ≊ H ≊ K ⇒ G ≊ K.
[Hint: Symmetry and transitivity follow from (3) and (2).]

(1.4.8) Definition. An isomorphism from G to itself is called an automorphism of G. The set of

all automorphisms of G is denoted Aut(G). It is a group under composition (see Exercise 1.4.9(2)).
(1.4.9) Exercises.
1) For g ∈ G, define φg : G → G by φg (x) = gx.
(a) Show that φg is an automorphism of G. It is called the conjugation by g (or the inner
automorphism corresponding to g).
(b) Show that the map g , φg is a homomorphism from G to Aut G.
2) Show that Aut(G) is a group under composition.
3) For n, k ∈ Z+, define φk : Zn → Zn by φk (x) = kx.
(a) Show that φk is a homomorphism.
(b) Show that φk is an automorphism of Zn if and only if gcd(k, n) = 1.
(c) For each φ ∈ Aut(Zn ), show there is some k ∈ Z+, such that φ = φk.
(d) Show that φk = φℓ if and only if k ≡ ℓ (mod n).
(1.4.10) Proposition (1st, 2nd, and 3rd Isomorphism Theorems).
1) If φ is a homomorphism from G to H, then G/ ker φ ≊ φ(G).
More precisely, an isomorphism φ : G/ ker φ → φ(G) can be obtained by defining
φ(g ker φ) = φ(g).
2) Suppose N ⊴ G and H is a subgroup of G. Then
(a) HN is a subgroup of G (where HN = { hn | h ∈ H, n ∈ N }),
(b) H ∩ N is a normal subgroup of H, and
(c) HN/N ≊ H/(H ∩ N).
3) Suppose H and N are normal subgroups of G, with N ⊆ H. Then H/N is a normal subgroup
of G/N, and (G/N)/(H/N) ≊ G/H.
Sketch of proof. (1) For g ∈ G, we have
φ(g ker φ) = φ(g) · φ(ker φ) = φ(g) · {1} = {φ(g)},
so φ is well defined. Also, if φ(g ker φ) = 1, then φ(g) = φ(g ker φ) = 1, which means g ∈ ker φ,
so g ker φ = ker φ is the trivial coset; therefore ker φ is trivial, so φ is one-to-one. Finally, for
h ∈ φ(G), there exists g ∈ G, such that φ(g) = h, so φ(g ker φ) = φ(g) = h; therefore φ is onto.
 1.4. HOMOMORPHISMS AND ISOMORPHISMS 15

(2a) It suffices to show that HN is closed under multiplication and inverses. Since N ⊴ G, we
have hN = Nh for all h ∈ H; therefore HN = NH. Hence,
(HN)(HN) = H(NH)N = H(HN)N = (HH)(NN) = HN,
so HN is closed under multiplication. Also,
(HN)−1 = N −1 H −1 = NH = HN,
so HN is closed under inverses.
(2b) For h ∈ H, we have h(H ∩ N)h−1 = (hHh−1 ) ∩ (hNh−1 ) = H ∩ N, so H ∩ N ⊴ H.
(2c) Let φ : G → G/N be the natural homomorphism with kernel N, and let φ′ be the restriction
of φ to H. Then ker φ′ = H ∩ ker φ = H ∩ N, so applying (1) to the homomorphism φ′ yields
HN/N = φ′ (H) ≊ H/ ker φ′ = H/(H ∩ N).
(3) Let φ : G → G/H be the natural homomorphism. The proof of (1) implies that there is
a well-defined surjective homomorphism φ : G/N → G/H, defined by φ(gN) = φ(g), and that
ker φ = H/N. Then applying (1) to the homomorphism φ yields
(G/N)/(H/N) = (G/N)/ ker φ ≊ φ(G/N) = G/H. □

(1.4.11) Remark. When you need to show that a quotient group G/N is isomorphic to some
group H, you should almost never try to directly construct an isomorphism from G/N to H.
Instead, Proposition 1.4.10(1) says that it suffices to find a homomorphism from G onto H whose
kernel is N. This is usually much easier.
(1.4.12) Proposition (Correspondence Theorem). Suppose N is a normal subgroup of G. Then there
is a one-to-one correspondence between the subgroups of G that contain N and the subgroups of
G/N. Namely, the subgroup of G/N corresponding to a subgroup H of G is H/N.
Furthermore, this remains a one-to-one correspondence when restricted to the normal subgroups
of G and G/N.
Sketch of proof. Let
• H be the collection of all subgroups of G that contain N,
• H be the collection of all subgroups of G/N, and
• φ : G → G/N be the natural homomorphism with kernel N.
Define:
• φ : H → H by φ(H) = φ(H) (= H/N), and
• φb : H → H by φb (K) = φ−1 (K).
It is straightforward to verify that φ and φ b are inverses of each other, so φ is a bijection.
For K ∈ H , it is also straightforward to verify that K ⊴ G/N a φ b (K) ⊴ G. □
 Chapter 2

Group Actions

§2.1. Definition and basic facts

(2.1.1) Definition. A (left) action of a group G on a set X is a function α : G × X → X that satisfies
the following two axioms. (We often write g ∗ x or gx or gx as shorthand for α(g, x).)
1) g(hx) = (gh)x for all g, h ∈ G and x ∈ X, and
2) 1x = x for all x ∈ X (where 1 is the identity element of G).
Here is another way to think about group actions:
(2.1.2) Lemma. Suppose G acts on X, and SX is the group of all permutations of X. For each g ∈ G,
define φg : X → X by φg (x) = gx. Then
1) φg ∈ SX , and
2) φ : G → SX is a homomorphism.
Idea of proof. (1) Verify that φg−1 is the inverse of φg:
( )
φg−1 φg (x) = φg−1 (gx) = g −1 (gx) = (g −1 g)x = 1x = x.
Any function with an inverse is a bijection, so this implies that φg is a bijection.
(2) Verify that φgh = φg φh: for x ∈ X, we have
( )
φgh (x) = (gh)x = g(hx) = φg (hx) = φg φh (x) = (φg φh )(x). □

(2.1.3) Exercise. Conversely, verify that every homomorphism φ : G → SX yields an action of G

on X, by defining gx = φg (x).
(2.1.4) Definition. Recall that every homomorphism φ has a kernel:
ker φ = { g ∈ G | φ(g) = 1 }.
The kernel of a group action is the kernel of the corresponding homomorphism:
{ g ∈ G | φg is the trivial permutation } = { g ∈ G | ∀x ∈ X, gx = x }.
(2.1.5) Example (regular action). G acts on itself by multiplication on the left:
for X = G, we may let g ∗ x = gx (multiplication in the group).
You probably saw this action in your undergraduate abstract algebra class, because it is used
in the proof of the following fundamental result:
(2.1.6) Basic fact (Cayley’s Theorem). Every finite group is isomorphic to a subgroup of Sn, for
some n. More precisely, we may take n = |G|.
17
18 2. GROUP ACTIONS

You should verify for yourself that this (and the following examples) satisfy the conditions to
be an action. (For example, the above example uses (1) the fact that groups are associative, and
(2) the definition of the identity element.)
(2.1.7) Examples. Every group has the following actions (in addition to the regular action that was
described in Example 2.1.5):
1) G acts trivially on any set X, by letting gx = x for all g ∈ G and x ∈ X.
2) G acts on itself by conjugation, letting gx = gxg −1. (On the right-hand side, the
multiplication is the group operation of G.)
3) For any subgroup H of G, recall that G/H is the set of left cosets of H in G:
G/H = { gH | g ∈ G}.
G acts on this set by left multiplication: g ∗ xH = gxH.
(2.1.8) Examples.
1) As in Lemma 2.1.2 above, let SX be the group of all permutations of X. Then SX acts on X
in a natural way. (The identity map is a homomorphism from G = SX to SX .) It also acts on
the power set P(X) (the set of all subsets of X), by σ A = { σ (a) | a ∈ A } for σ ∈ SX and
A ⊆ X.
2) More generally, if G acts on X, then G also acts on P(X), by gA = { ga | a ∈ A } for A ⊆ X.
3) Let H be a group. If G acts on X, and we have a homomorphism φ : H → G, then an action
of H on X can be defined by letting hx = φ(h) x.
4) Any action of G on X can be restricted to any subgroup of G. More precisely, if G acts on X,
and H is a subgroup of G, then an action of H on X is obtained by defining h ∗ x = hx.
(2.1.9) Other conventions. Some authors use right actions, instead of left actions, so α : X × G → X,
and the axioms are: (xg)h = x(gh) and x1 = x. To compensate for this, the action by conjugation
is defined by x g = g −1 xg.

§2.2. Orbits and stabilizers

(2.2.1) Assumption. In this section, we assume that G acts on X.
(2.2.2) Definitions. Let x ∈ X.
1) The orbit of x under G is Gx = { gx | g ∈ G }.
2) The stabilizer of x in G is StabG (x) = { g ∈ G | gx = x }.
3) If Gx = X for some (or, equivalently, all) x ∈ X, then we say that the action is transitive.
(2.2.3) Exercises.
1) Show that G is transitive on X if and only if for all x, y ∈ X, there exists g ∈ G, such that
gx = y.
2) Show that the orbits form a partition of X. That is, every element of X is in a unique orbit
(and no orbit is the empty set). In other words, show that
(a) the union of all of the orbits is X,
(b) any two different orbits are disjoint, and
(c) no orbit is the empty set.
3) Show StabG (x) is a subgroup of G, for all x ∈ X.
4) For g ∈ G and x ∈ X, show StabG (gx) = g StabG (x) g −1. (This implies that if x and y are
in the same orbit, then StabG (x) is conjugate to StabG (y), so the two stabilizers have the
same order.)
5) Suppose N ⊴ G, and N ⊆ StabG (x), for some x ∈ X. Show that if G is transitive on X, then
N is contained in the kernel of the action.
 2.2. ORBITS AND STABILIZERS 19

6) Suppose
(a) x ∈ X, g, h ∈ G, and
(b) K is a subgroup of G that contains StabG (x).
Let
Y = Kx = { kx | k ∈ K }, gY = { gy | y ∈ Y }, hY = { hy | y ∈ Y }.
Show that if gY is not disjoint from hY, then gY = hY.
(2.2.4) Examples.
1) For the regular action of G (by left-multiplication on itself):
• The action is transitive. (For x, y ∈ G, if we let g = yx −1, then y = gx.)
• StabG (x) is trivial. (If gx = x, then g = (gx)x −1 = xx −1 = 1.)
In this case, the orbit is large (all of G) and the stabilizer is very small (trivial).
2) At the other extreme, for the trivial action of G on X:
• Each orbit is only a single point. (If gx = y, then x = y.)
• StabG (x) = G. (We have gx = x for all g.)
In this case, every orbit is small (a single point), and the stabilizers are large (all of G).
3) Let A ⊆ X, and let k = #A. For the action of SX on P(X):
• We have SX A = { B ∈ P(X) | #B = k }. So the number of orbits is exactly 1 + #X
(because k can be any number from 0 to #X).
• StabSX (A) ≊ SA × SX∖A.
(2.2.5) Exercise. For any subgroup H of G, show the action of G on G/H (by left multiplication) is
transitive, and that StabG (gH) = gHg −1, for all g ∈ G.
The following important result shows that points with large stabilizers always have small
orbits, and points with small stabilizers have large orbits. More precisely, the cardinality of the
orbit is the index of the stabilizer:
(2.2.6) Theorem (Orbit-Stabilizer Theorem). Assume G acts on X, and x ∈ X. Then
|Gx| = |G : StabG (x)|.
Proof. For convenience, let H = StabG (x), so |G : StabG (x)| = |G/H|. Define f : G/H → Gx by
f (gH) = gx. Then f is:
• onto: Given y ∈ Gx, there exists g ∈ G, such that y = gx = f (gH).
• one-to-one: If f (g1 H) = f (g2 H), then g1 x = g2 x, so (multiplying both sides on the left
by g2−1, we have g2−1 g1 x = x. This means g2−1 g1 ∈ StabG (x) = H, so g1 H = g2 H.
• well-defined: If g1 H = g2 H, then g2−1 g1 ∈ H = StabG (x), so
f (g1 H) = g1 x = g2 (g2−1 g1 )x = g2 x = f (g2 H).
Thus, f is a bijection from G/H to Gx, so the two sets have the same cardinality. □

Here is another way of saying the same thing:

(2.2.7) Corollary. |StabG (x)| · |Gx| = |G|.
(2.2.8) Remark. Lagrange’s Theorem tells us |StabG (x)| is a divisor of G (if G is finite). The corollary
tells us that |Gx| is also a divisor of G.
(2.2.9) Example. The group Rot(cube) of all rotations of a cube acts transitively on the six faces
of the cube. (If you pick up a cube, you can set it down into the exactly the same space it was in
before, with any face you choose on the bottom.) And there are exactly four rotations of the cube
that keep a given face fixed (setwise). So the Orbit-Stabilizer Theorem implies that the order of
Rot(cube) is 6 × 4 = 24.
(2.2.10) Exercises.
1) Show that Rot(cube) is isomorphic to the symmetric group S4 on four symbols.
[Hint: Rot(cube) permutes the four diagonals of the cube.]
20 2. GROUP ACTIONS

2) Show that if H and K are subgroups of G, and G is finite, then

|HK| = |H| |K|/|H ∩ K|.
[Hint: Restrict the action of G on G/K to H, and apply the Orbit-Stabilizer Theorem. We have StabH (K) = H ∩ K,
and the cardinality of the H-orbit of K is |HK|/|K|.]
3) Assume
(a) p is a prime number that divides |G|, but does not divide |X|, and
(b) p r is the largest power of p that divides |G|.
Show there exists x ∈ X, such that |StabG (x)| is divisible by p r .
(2.2.11) Application. Let us apply the Orbit-Stabilizer Theorem to the action of G on itself by
conjugation: gx = gxg −1. Each orbit of this action is called a conjugacy class in G. Since the
orbits (or conjugacy classes) partition G, we have
( )
∑ number of conjugacy ∑
|G| = |C| = + |C|.
classes of cardinality 1
conjugacy classes C conjugacy classes C
with |C| > 1
Note that:
• The conjugacy class of x has cardinality 1 if and only if x commutes with every element
of G, which means x ∈ Z(G) (the centre of G). So the number of conjugacy classes of
cardinality 1 is |Z(G)|.
• The Orbit-Stabilizer Theorem tells us |C| = |G : StabG (x)| for x ∈ C.
• We have g ∈ StabG (x) if and only if g commutes with x, which means g ∈ CG (x) (the
centralizer of x).
Therefore, if we choose an element xi from each conjugacy class of cardinality > 1, then
∑
|G| = |Z(G)| + |G : CG (xi )|. (2.2.12)
i
This is the class equation. It is an important tool in the theory of finite groups.
We can also now complete some unfinished business from Section 1.2:
Proof of Burnside’s Counting Lemma (1.2.3). This proof employs a standard technique: count
the elements of a set in two different ways. Let
F = { (g, x) ∈ G × X | gx = x }.
For each g ∈ G, let f (g) be the number of fixed points of g, so
f (g) = #{ x ∈ X | (g, x) ∈ F }.
Therefore ∑
#F = f (g).
g∈G
Also, for any x ∈ X, we have StabG (x) = { g ∈ G | (g, x) ∈ F }, so
∑
#F = |StabG (x)|.
x∈X
Therefore
∑ ∑
f (g) = |StabG (x)|. (2.2.13)
g∈G x∈X

Recall that when x and y are in the same orbit, we have |StabG (x)| = |StabG (y)| (see
Exercise 2.2.3(4)). Hence, for each x ∈ X, we have
∑ ∑
|StabG (y)| = |StabG (x)| = |Gx| · |StabG (x)| = |G| (2.2.14)
y∈Gx y∈Gx
(by the Orbit-Stabilizer Theorem). We can break up the right-hand-side of (2.2.13) into sums over
orbits, and (2.2.14) shows that the sum over each orbit is |G|. So we conclude that the right-hand
side of ∑(2.2.13) is equal to |G| times the number of orbits in X. Dividing by |G|, we find that
(1/|G|) g∈G f (g) equals the number of orbits, as desired. □
 2.3. SYLOW THEOREMS 21

§2.3. Sylow Theorems

(2.3.1) Definitions. Suppose G is finite, p is a prime number, and P is a subgroup of G.
1) G is a p-group if |G| is a power of p. (That is |G| = p k, for some k ∈ N.)
2) P is a p-subgroup of G if |P | is a power of p. (In other words, a p-subgroup is a subgroup
that also happens to be a p-group.)
3) P is a Sylow p-subgroup of G if |P | is the largest power of p that divides |G|.
4) Sylp (G) is the set of all Sylow p-subgroups of G.
5) P is a Sylow subgroup of G if it is a Sylow p-subgroup of G for some prime p.
(2.3.2) Example. Suppose |G| = 600 = 23 · 3 · 52. Then the Sylow p-subgroups of G are the
subgroups of order:
23 if p = 2, 3 if p = 3, 52 if p = 5, 1 if p > 5.
So the Sylow subgroups of G are the subgroups of order 1, 3, 8, or 25.
(2.3.3) Remark. By Lagrange’s Theorem, we know that if P is any subgroup of G, then |P | is a
divisor of |G|. Thus, to say that P is a Sylow p-subgroup means that it is a p-subgroup of the
largest possible order that is compatible with Lagrange’s Theorem — there could not possibly be
a p-subgroup of larger order.
In this section, we establish the following fundamental result that is often stated, but usually
not proved, in the first semester undergraduate abstract algebra.
(2.3.4) Theorem (Sylow Theorems). Let G be a finite group and let p be a prime number. Then:
1) (existence) G has a Sylow p-subgroup.
2) (conjugacy) Any two Sylow p-subgroups of G are conjugate.
3) (development) Every p-subgroup of G is contained in a Sylow p-subgroup of G.
4) The number of Sylow p-subgroups of G is a divisor of |G|, and is congruent to 1, modulo p.
Our proof of the existence of Sylow p-subgroups will use the following fact from elementary
number theory.
( r )
(2.3.5) Lemma. Assume p is prime, r ∈ N, and m ∈ Z+, such that p ∤ m. Then p ∤ ppm , where the
( ) r

n n!
notation k denotes the binomial coefficient k! (n−k)! .

We sketch two of the numerous different proofs of this fact.

( ) ( )
pn n
Proof 1. By induction on r, it suffices to show pk ≡ k (mod p) for n, k ∈ N. Since p is
( )
prime, it is well known (and easy to prove from the fact that p | pk for 1 ≤ k ≤ p − 1) that
(a + b)p ≡ ap + bp (mod p). So
(1 + x)pn ≡ (1 + x p )n (mod p).
( )
By the binomial theorem, the coefficient of x pk on the left-hand side is pn , but the coefficient on
( ) pk
n
the right-hand side is k . So these two binomial coefficients must be congruent, modulo p. □

Proof 2. We have
( ) p −1 r
r
pr m (p r m)! ∏ p m−i
= = .
pr (p )! (p m − p )!
r r r
i=0
pr − i
In each of the factors in this product, elementary number theory reveals that the largest power
of p that divides the numerator is exactly the same as the largest power of p that divides the
denominator. That is, all occurrences of p cancel, so no factor in this product is divisible by p.
Hence, the product is not divisible by p. □
22 2. GROUP ACTIONS

(2.3.6) Remark. Actually, something much stronger than Lemma 2.3.5 is true. Namely, suppose
we write n and r in base p (where p is prime):
n = nr p r + nr −1 p r −1 + · · · + n1 p + n0 and k = kr p r + kr −1 p r −1 + · · · + k1 p + k0,
with ni and ki in {0, 1, . . . , p − 1}. Then
( ) ( )( ) ( )
n nr nr −1 n0
≡ ··· (mod p).
k kr kr −1 k0
This can be established by the method of Proof 1.
Proof of Sylow’s Existence Theorem. Let p r be the largest power of p that divides |G|. We wish
to show that G has a subgroup of order p r .
To accomplish this, let
X = { A ⊆ G | |A| = p r }
r
be the collection of all subsets (not just subgroups) of G that( have) cardinality p . Then G acts on X
|G|
by left multiplication: g · A = { ga | a ∈ A }. Since |X| = pr , and |G|/p r is not divisible by p,
Lemma 2.3.5 tells us that |X| is not divisible by p, so there must an orbit G · A whose cardinality
is not divisible by p. Since
|G|
|StabG (A)| = ,
|G · A|
we see that p r is a divisor of |StabG (A)|.
On the other hand, for any a ∈ A, we have StabG (A)a ⊆ StabG (A) A = A, so
|StabG (A)| ≤ |Aa−1 | = |A| = p r .
Therefore, we must have |StabG (A)| = p r , so StabG (A) is a Sylow p-subgroup of G. □

The following lemma is a crucial tool in the remaining parts of the proof.
(2.3.7) Definition. Suppose G acts on X, and x ∈ X. We say x is a fixed point of the action if
gx = x for all g ∈ G. In other words, |Gx| = 1.
(2.3.8) Lemma. For any action of a p-group P on a finite set X, the number of fixed points is
congruent to |X|, modulo p.
Proof. Write X = F ⊔ Y (disjoint union), where F is the set of fixed points, and Y is the union of
the orbits of cardinality > 1. The Orbit-Stabilizer Theorem tells us that the cardinality of every
orbit is a divisor of |P |, and is therefore a power of p. Hence, the cardinality of each orbit in Y is
multiple of p (since 1 is the only power of p that is not divisible by p). Since these orbits form a
partition of Y, we conclude that |Y | is a sum of multiples of p, and is therefore itself a multiple
of p. So |F | = |X| − |Y | ≡ |X| (mod p). □
Proof of Sylow’s Conjugacy Theorem. Let P , Q ∈ Sylp (G). We wish to show that Q is conjugate
to P. To do this, we begin by repeating part of the proof of Sylow’s Development Theorem.
The group G acts on G/P by multiplication on the left: g ∗ (hP ) = ghP (see Example 2.1.7(3)).
Restrict this to an action of Q on G/P. Since Q is a p-group and |G/P | = |G|/|P | is not divisible
by p (since P is a Sylow p-subgroup), we see from Lemma 2.3.8 that Q must have at least one fixed
point gP in G/P. This means Q ⊆ StabG (gP ) = gP.
However, since P and Q are Sylow p-subgroups, we know that |P | = |Q|. Also, since
conjugate subgroups have the same order (see Exercise 1.4.6(2)), we have |gP | = |P |. Therefore
|Q| = |P | = |gP |. Since Q ⊆ gP (and the groups are finite), this implies Q = gP. Thus, we have
shown that Q is a conjugate of P, as desired. □
Proof of Sylow’s Development Theorem. Given any p-subgroup Q of G, we wish to show that Q is
contained in some Sylow p-subgroup. By Sylow’s Existence Theorem, we may fix some P ∈ Sylp (G).
The group G acts on G/P by multiplication on the left: g ∗ (hP ) = ghP (see Example 2.1.7(3)).
Restrict this to an action of Q on G/P. Since Q is a p-group and |G/P | = |G|/|P | is not divisible
by p (since P is a Sylow p-subgroup), we see from Lemma 2.3.8 that Q must have at least one fixed
 2.3. SYLOW THEOREMS 23

point gP in G/P. This means Q ⊆ StabG (gP ) = gP. Also, since conjugate subgroups have the same
order (see Exercise 1.4.6(2)), and P is a Sylow p-subgroup, we know that gP ∈ Sylp (G). Therefore,
we have shown that Q is contained in the Sylow p-subgroup gP. □

Our proof of the final part will use the following observation.

(2.3.9) Exercise. Let P be a Sylow p-subgroup of G and suppose Q is a p-subgroup of NG (P ) (the

normalizer of P). Show Q ⊆ P.
[Hint: P Q is a p-subgroup of G that contains P.]

(2.3.10) Exercise. Let P and Q be Sylow p-subgroups of G. If P ⊆ NG (Q), then P = Q.

[Hint: P Q is a p-subgroup of G that contains P.]

Proof of part (4) of Sylow’s Theorem. Fix some P ∈ Sylp (G). Since conjugate subgroups
have the same order, G acts on the set Sylp (G) by conjugation. Furthermore, Sylow’s Conjugacy
Theorem tells us that this action is transitive. Therefore, the Orbit-Stabilizer Theorem tells us
|Sylp (G)| = |G : NG (P )|. This is a divisor of |G|.
Restricting this conjugation action to P yields an action of P by conjugation on Sylp (G).
From Exercise 2.3.9, we see that the only fixed point is P. So Lemma 2.3.8 tells us that
|Sylp (G)| ≡ 1 (mod p). □

(2.3.11) Exercises. Assume G is finite, p is a prime number, and P ∈ Sylp (G).

1) Show the following are equivalent:
(a) P ⊴ G.
(b) P is the unique Sylow p-subgroup of G.
2) Let n be the number of Sylow p-subgroups of G. Show G has a subgroup of index n.
3) Suppose N ⊴ G. Show :
(a) P ∩ N ∈ Sylp (N).
(b) P N/N ∈ Sylp (G/N).

(2.3.12) Exercises.
1) Show that if G/Z(G) is cyclic, then G is abelian.
2) Let p be a prime number. Show that every group of order p 2 is abelian.
[Hint: Groups of prime order are cyclic. Use Exercise 1.]

3) Assume p and q are prime numbers, with p > q. Show that if p ̸≡ 1 (mod q), then every
group of order pq is abelian.
[Hint: Let P be a Sylow p-subgroup of G. Since 1 is the only divisor of |G| that is ≡ 1 (mod p), we know P ⊴ G, so
G acts on P by conjugation. However, |Aut(P )| = p − 1 (see Exercise 1.4.9(3)), which is relatively prime to |G|, so
the action must be trivial, which means P ⊆ Z(G).]

4) Assume p and q are prime numbers, with p > q. Show that every group of order pq is
isomorphic to the semidirect product Zp ⋊k Zq, for some k ∈ Z+. (See Exercise 1.1.8 for the
definition of the semidirect product.)
5) Recall that A4 is the alternating group of degree 4. (This has order 12, because it is a
subgroup of index 2 in the symmetric group S4, which has order 4! = 24.)
(a) How many Sylow 3-subgroups does A4 have?
(b) How many elements of order 3 does A4 have?
(c) How many Sylow 2-subgroups does A4 have?
6) Show that if |P | = p n, then P has a normal subgroup of order p k, for 0 ≤ k ≤ n.
[Hint: Exercise 2.3.13(1), induction on |P |, and the Correspondence Theorem.]

7) Suppose Q is a p-subgroup of G, and Q ⊴ G. Show that Q is contained in every Sylow

p-subgroup of G.
24 2. GROUP ACTIONS

(2.3.13) Exercises. Use Lemma 2.3.8 to prove:

1) Every nontrivial p-group has nontrivial centre. That is, if P is a p-group, and P ≠ {1}, then
Z(P ) ≠ {1}.
[Hint: Let P act on itself by conjugation (gx = gxg −1). Fixed points are precisely the elements of Z(P ), and we
know that 1 ∈ Z(P ).]
2) (Cauchy’s Theorem) If |G| is divisible by p, then G has an element of order p.
[Hint: Let
X = { (g1 , g2 , . . . , gp ) ∈ Gp | g1 g2 · · · gp = 1 }.
Then Zp acts on X by cyclic permutations:
k ∗ (g1 , g2 , . . . , gp ) = (gk+1 , gk+2 , . . . , gk+p ),
where the subscripts are read modulo p. Since |X| = |G|p−1 is divisible by p, the number of fixed points must
be divisible by p. One fixed point is (1, 1, . . . , 1). Any other fixed point is of the form (g, g, . . . , g), where g is an
element of order p.]
3) (Fermat’s Little Theorem) If p is prime and k ∈ Z, then kp ≡ k (mod p).
[Hint: Assume, for simplicity, that k ∈ Z+. Then the group Zp acts by rotations on the set of all possible colourings
of the vertices of the regular p-gon with k colours of paint.]

It is also instructive to have alternate proofs of the results in Exercise 2.3.13:

(2.3.14) Exercises.
1) Derive Exercise 2.3.13(1) from the class equation (2.2.12).
[Hint: If x ∉ Z(P ), then |P : CP (x)| is divisible by p (why?).]
2) Derive Cauchy’s Theorem 2.3.13(2) from Sylow’s Theorem.
[Hint: Let P be a Sylow p-subgroup of G, and let g be any nontrivial element of P. Lagrange’s Theorem tells us
that |g| = p k for some k. Then g k−1 is an element of order p.]
3) Derive Fermat’s Little Theorem 2.3.13(3) from Lagrange’s Theorem.
[Hint: Assume k ̸≡ 0 (mod p), so k ∈ Z× ×
p , the group of units of the ring Zp . Since |Zp | = p − 1, Lagrange’s Theorem
implies k p−1 ≡ 1 (mod p).]
 Index

abelian, 2 kernel of a homomorphism, 17

action of a group, 17
automorphism, 14 Lagrange’s Theorem, 12
left coset, 12
bijection, 6
nontrivial subgroup, 10
binomial coefficient, 21
normal subgroup, 11
cardinality, 1 normalizer, 10
Cauchy’s Theorem, 24 normalizes, 11
Cayley’s Theorem, 17
one-to-one, 5
centralizer, 10
onto, 6
centralizes, 2
orbit, 18
centre, 10
Orbit-Stabilizer Theorem, 19
class equation, 20
order, 2
commutative group, 2
commute, 2 p-group, 21
composition, 6 p-subgroup, 21
conjugacy class, 20 partition, 12, 18
conjugate, 11 permutation, 6
conjugation by g, 14 power set, 18
Correspondence Theorem, 15 proper subgroup, 10
cycle, 6
cyclic, 10 quotient, 12
cyclic subgroup, 10
regular action, 17
dihedral group, 5 restricting an action to a subgroup, 18
direct product, 3 right coset, 12
disjoint, 6
same cardinality, 6
Fermat’s Little Theorem, 24 semidirect product, 3
fixed point, 7, 22 stabilizer, 18
subgroup, 10
generates, 11 subgroup generated by S, 11
generating set, 11 surjective, 6
group, 1 Sylow p-subgroup, 21
Sylow subgroup, 21
homomorphism, 13 Sylow Theorems, 21
Sylp (G), 21
identity map, 7 symmetric group, 6
image, 6 symmetric group on n letters, 6
index of a subgroup, 12 symmetry, 3
injective, 5
inner automorphism, 14 transitive action, 18
integers modulo n, 3 trivial action, 18
isomorphic, 13 trivial subgroup, 10
isomorphism, 13
Isomorphism Theorems (1st, 2nd, and 3rd), 14

k-cycle, 6
kernel, 13
kernel of a group action, 17

25